TITLE

METHODS OF RECOVERING CATALYSTS

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to the U.S. Provisional Patent

Application No. 62/281,440, filed January 21, 2016, the contents of the entirety of which are incorporated by this reference.

TECHNICAL FIELD

[0002] The present invention relates generally to catalysts. The present disclosure is further directed to methods of separating a catalyst from a product mixture. The present disclosure is also directed to methods of recovering a catalyst from an epoxidized vegetable oil product mixture.

BACKGROUND OF THE INVENTION

[0003] Methods of epoxidation of vegetable oils are well known. For example US Patent Application US 2006/0020062 Al to Bloom discloses that epoxidation of soybean and linseed oils is well known in the art, is performed on a commercial scale, and commonly is achieved by using a strong catalyst, oxidant, and carboxylic acid. ¹ Epoxidized vegetable oils have a wide range of utility, including uses as plasticizers and stabilizers in certain polymers. ²

[0004] Heterogeneous systems for epoxidation are well known. For example, "Metal-catalyzed Epoxidations of Alkenes with Hydrogen Peroxide" discloses several typical heterogeneous systems, such as mineral-type catalysts including zeolites and hydrotalcites. ³ Additionally, heterogeneous systems may be constructed by attaching homogeneous catalysts to solid supports. ⁴ While heterogeneous systems have the advantage of easy catalyst recovery, these systems suffer from drawbacks such as limitations in the kinds of epoxides which can be produced and reduced activity relative to homogeneous catalysts. ⁵

[0005] Homogeneous systems for epoxidation are also well known. For example, "Metal-catalyzed Epoxidations of Alkenes with Hydrogen Peroxide" discloses several kinds of soluble metal oxides, including the Venturello epoxidation catalyst and the Noyori epoxidation system. ⁶ The epoxidation of oils catalyzed by organometallic compounds is advantageous, for example being environmentally benign and efficient in both activity and selectivity. ⁷ However, concerns with tungsten-based systems include cost and product-impurities. ⁸ Additionally, catalyst recovery is more difficult relative to heterogeneous systems for epoxidation.

[0006] Overall, it would be especially advantageous to develop methods for recovering catalysts used in homogeneous systems for epoxidation.

SUMMARY OF THE INVENTION

² US 2006/0020062 Al, paragraph [0005] ,

³ Lane, Benjamin; Burgess, Kevin. Metal-Catalyzed Epoxidations of Alkenes with Hydrogen Peroxide. Chem. Rev. 2003, 103 (7), 2458.

⁴ Lane, Benjamin; Burgess, Kevin. Metal-Catalyzed Epoxidations of Alkenes with Hydrogen Peroxide. Chem. Rev. 2003, 103 (7), 2459.

⁵ Lane, Benjamin; Burgess, Kevin. Metal-Catalyzed Epoxidations of Alkenes with Hydrogen Peroxide. Chem. Rev. 2003, 103 (7), 2458-59. ,

⁶ Lane, Benjamin; Burgess, Kevin. Metal-Catalyzed Epoxidations of Alkenes with Hydrogen Peroxide. Chem. Rev. 2003, 103 (7), 2459-60.

⁷ Jiang, Pingping; Chen, Min; Dong, Yuming; Lu, Yun; Ye, Xia; Zhang, Weije. Novel Two-Phase Catalysis with Organometallic Compounds for Epoxidation of Vegetable Oils by Hydrogen Peroxide. J. Am. Oil Chem. Soc. 2010, 87, 90.

⁸ Lane, Benjamin; Burgess, Kevin. Metal-Catalyzed Epoxidations of Alkenes with Hydrogen Peroxide. Chem. Rev. 2003, 103 (7), 2460. [0007] The present invention relates in one aspect to a method of separating a catalyst from a product mixture comprising diluting the product mixture with a solvent, passing the diluted product mixture through a first ion exchange resin column, and passing the diluted product mixture through a second ion exchange resin column, thus retaining the catalyst in either the first ion exchange resin column, the second ion exchange resin column, or combinations of any thereof.

[0008] The present invention relates in another aspect to a method of separating a catalyst from a product mixture comprising diluting the product mixture with a solvent, contacting the diluted product mixture with a first ion exchange resin, and contacting the diluted product mixture with a second ion exchange resin, thus retaining the catalyst in either the first ion exchange resin, the second ion exchange resin, or combinations of any thereof.

[0009] The present invention relates in yet another aspect to a method of recovering a catalyst from an epoxidized vegetable oil product mixture comprising diluting the epoxidized vegetable oil product mixture with an organic solvent, wherein the epoxidized vegetable oil product mixture comprises the catalyst, eluting the diluted epoxidized vegetable oil product mixture through a double bed resin column, wherein the double bed resin column is packed with an anion exchange resin and a cation exchange resin, and eluting a base through the double bed resin column, thus collecting the catalyst.

BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 shows a scheme for catalyst retention and recovery using a cation exchange resin followed by an anion exchange resin.

[0011] Figure 2 shows a scheme for catalyst retention and recovery using an anion exchange resin followed by a cation exchange resin.

[0012] Figure 3 shows a scheme for retention and removal of a catalyst using a single bed resin column.

[0013] Figure 4 shows results for catalyst separation using cation exchange resin followed by anion exchange resin in a mixed bed resin system, wherein the cation exchange resin is packed on top of the anion exchange resin within a single column. [0014] Figure 5 shows results for catalyst separation using cation exchange resin followed by anion exchange resin in a single bed resin system.

[0015] Figure 6 shows a scheme for removal of Venturello catalyst components using anion exchange resin.

[0016] Figure 7 shows a scheme for calculation of catalyst absorbed/retained in resin using gravimetric method/analysis.

[0017] Figure 8 shows a breakthrough curve for absorption/retention of catalyst using Purolite® Al 11 S (-NR2FLO ^" form) anion exchange resin.

[0018] Figure 9 shows results for recovered tungsten from anion exchange resin bed (Purolite® Al 11 S, -N ⁺ HR ₂ C1- form).

[0019] Figure 10 shows results for recovered phosphorous from anion exchange resin bed (Purolite® Al 11 S, -N ⁺ HR2C1 ^" form).

[0020] Figure 11 shows results for recovery of tungsten and phosphorous from anion exchange resin (Purolite® Al 11 S, -N ⁺ HR2C1 ^" form).

[0021] Figure 12 shows results for recovery of tungsten and phosphorous from anion exchange resin (Purolite® Al 11 S, -N ⁺ HR2C1 ^" form).

[0022] Figure 13 shows extent of tungsten and phosphorous recovery from anion exchange resin (Purolite® Al 11 S, -N ⁺ HR2C1 ^" form).

[0023] Figure 14 shows a scheme for conversion of -N ⁺ HR2C1 ^" form resin to - N ⁺ HR ₂ OH- form resin.

[0024] Figure 15 shows an exemplary resin bed set-up.

[0025] Figure 16 shows additional views of an exemplary resin bed set-up.

[0026] Figure 17 shows an additional view of an exemplary resin bed set-up.

[0027] Figure 18 shows an additional view of an exemplary resin bed set-up.

[0028] Figure 19 shows results of a calibration of flow rate using a Masterflex

L/S pump.

[0029] Figure 20 shows results of a calibration of flow rate using a Masterflex L/S pump.

DETAILED DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS OF THE

INVENTION [0030] In an illustrative embodiment, a method of the present invention includes separating a catalyst from a product mixture comprising diluting the product mixture with a solvent, passing the diluted product mixture through a first ion exchange resin column, and passing the diluted product mixture through a second ion exchange resin column, thus retaining the catalyst in either the first ion exchange resin column, the second ion exchange resin column, or combinations of any thereof.

[0031] In another illustrative embodiment, a method of the present invention includes separating a catalyst from a product mixture comprising diluting the product mixture with a solvent, contacting the diluted product mixture with a first ion exchange resin, and contacting the diluted product mixture with a second ion exchange resin, thus retaining the catalyst in either the first ion exchange resin, the second ion exchange resin, or combinations of any thereof.

[0032] In yet another illustrative embodiment, a method of the present invention includes recovering a catalyst from an epoxidized vegetable oil product mixture comprising diluting the epoxidized vegetable oil product mixture with an organic solvent, wherein the epoxidized vegetable oil product mixture comprises the catalyst, eluting the diluted epoxidized vegetable oil product mixture through a double bed resin column, wherein the double bed resin column is packed with an anion exchange resin and a cation exchange resin, and eluting a base through the double bed resin column, thus collecting the catalyst.

[0033] In a further embodiment, the product mixture comprises an epoxidized vegetable oil. In one embodiment, the product mixture consists of a substance selected from the group consisting of an epoxidized vegetable oil, a catalyst, and combinations of any thereof.

[0034] The present invention contemplates many catalysts, including a catalyst comprising tungsten and phosphorous, as well as a catalyst comprising a Venturello catalyst.

[0035] The present invention contemplates many solvents, including an organic solvent, as well as a solvent selected from the group consisting of ethanol, toluene, and combinations of any thereof. In one embodiment, the solvent is combined with the product mixture in a ratio of 3 : 1 volume/volume. [0036] In a further embodiment, the first ion exchange resin column is a cation exchange resin column and the second ion exchange resin column is an anion exchange resin column.

[0037] In a further embodiment, the first ion exchange resin is an anion exchange resin and the second ion exchange resin is a cation exchange resin.

[0038] In a further embodiment, the first ion exchange resin is a cation exchange resin and the second ion exchange resin is an anion exchange resin.

[0039] In one embodiment, the anion exchange resin comprises an N+R.2HC1 ^" group. In another embodiment, the anion exchange resin comprises Lewatit® S 4528 resin.

[0040] In one embodiment, the cation exchange resin comprises Lewatit® TP 207 resin.

[0041] In a further embodiment, the contacting the diluted product with the first ion exchange resin is achieved by a method selected from the group consisting of shaking the first ion exchange resin with the diluted product mixture, stirring the first ion exchange resin with the diluted product mixture, and passing the diluted product mixture through a column packed with the first ion exchange resin.

[0042] In a further embodiment, the contacting the diluted product mixture with the second ion exchange resin is achieved by a method selected from the group consisting on shaking the second ion exchange resin with the dilute product mixture, stirring the second ion exchange resin with the diluted product mixture, and passing the diluted product mixture through a column packed with the second ion exchange resin.

[0043] In a further embodiment, the passing the diluted product mixture through a column packed with the first ion exchange resin and the passing the diluted product mixture through a column packed with the second ion exchange resin is at a rate of from about 2 bed volumes (BV)/hour to about 4 bed volumes (BV)/hour.

[0044] In a further embodiment, the first ion exchange resin and the second ion exchange resin are contained in two, successive single bed resin columns.

[0045] In a further embodiment, the first ion exchange resin and the second ion exchange resin are contained in a double bed resin column.

[0046] In a further embodiment, the catalyst is recovered from the first ion exchange resin, the second ion exchange resin, or combinations of any thereof. In one embodiment, the recovering the catalyst from the first ion exchange resin, the second ion exchange resin, or combinations of any thereof comprises contacting the first ion exchange resin, the second ion exchange resin, or combinations of any thereof with a base.

[0047] The present invention contemplates many bases, including a base selected from the group consisting of sodium hydroxide, ammonium hydroxide, sodium bicarbonate, ammonium bicarbonate, and combinations of any thereof. In one embodiment, the base has a concentration of at least 1 Normal. In another embodiment, the base has a concentration of between about 1 Normal and about 8 Normal.

[0048] In a further embodiment, the first ion exchange resin and the second ion exchange resin are present in equal amounts. In one embodiment, the first ion exchange resin and the second ion exchange resin ranges from a total combined weight of about 4 grams to about 80 grams.

[0049] In a further embodiment, the anion exchange resin is packed on top of the cation exchange resin.

[0050] In a further embodiment, the anion exchange resin and the cation exchange resin are present in equal amounts. In one embodiment, the anion exchange resin and the cation exchange resin are present in a total amount of from about 4 grams to about 80 grams.

[0051] Referring now to the drawings, various schema are provided for carrying out the above illustrative embodiments. Thus, in Figure 1, a general scheme for catalyst retention and recovery using a cation exchange resin followed by an anion exchange resin is shown. The Venturello catalyst can be viewed as a salt of an anion (of a soft Lewis base) and a cation (of a soft Lewis acid). The Venturello catalyst is composed of a tri-negatively charged peroxophosphotungsten moiety

[P04{WO(02)2}4] ^{3 "} and a mono-positively charged suitable phase transfer cation. In certain preferred embodiments, the phase transfer cation is the methyltrioctylammonium cation [ {(C8Hn)3N(CH3)} ⁺ ] . Therefore, the cation exchange resin may include various functional groups including monocarboxylic acid and iminodiacetic acid as weak acid cations, as well as sulfonic acid and phosphonic acid as strong acid cations. The functional group of the weak anion exchange resin may be a primary, secondary, or tertiary amine group. The resin matrix is preferably composed of styrene divinyl benzene copolymer. When a pure catalyst solution or an organic phase of an epoxidation reaction mixture comprising soybean oil, soybean oil epoxide, an insignificant amount of by-products (from about 1% to about 3%), and primarily diols and Venturello catalyst is passed through the cation exchange resin first, a proton is exchanged with the phase transfer cation of the catalyst, giving rise to an ester reacting with the acid group. A pronated form of the Venturello catalyst which comes from the cation exchange resin bed is then passed through the anion exchange resin bed where it forms a salt upon reacting with the amine group. (Figure 1, Steps 1 and 2). In certain preferred embodiments, the cation exchange resin comprises a group represented by X in the diagram, wherein X is preferably an anion of acetic acid, iminodiacetic acid, sulfonic acid, or phosphonic acid. In certain preferred embodiments, the anion exchange resin comprises a group represented by R' in the scheme, wherein R' is preferably an H group, a CH3 group, or a CH2CH group. In certain preferred embodiments, for catalyst retention, the cation exchange resin comprises Lewatit® TP 207 and the anion exchange resin comprises Lewatit® S 4528. For catalyst recovery, a solution of any organic acid, including CH3COOH, HCOOH, (COOH) ₂ , H0 ₂ CCH(OH)CH(OH)C0 ₂ H, and in a suitable organic solvent, including toluene, heptane, acetonitrile, and ethyl acetate, may be passed through the cation exchange resin to recover the phase transfer reagent, represented as RA in Figure 1. (Figure 1, Step 3). In certain preferred embodiments, RA represents Aliquat® 336 (N-methyl-N,N,N-trioctyloctan-l- ammonium chloride) as a salt of an anionic part of the acid used. In certain

embodiments, an inorganic acid solution in an organic solvent may be used instead of an organic acid in a suitable organic solvent, with preferred inorganic acids including HC1, ΗΝΟ3, H2SO4, and H2SO3. In certain preferred embodiments, the solvent is ethanol. In order to recover the anionic portion of the Venturello catalyst (the portion containing tungsten and phosphorous), an aqueous solution of a suitable salt is passed through the anion exchange resin. (Figure 1, Step 4). The suitable salt is represented as ΜΆ', wherein M' is preferably K ⁺ , Na ⁺ , NH4 ⁺ , Ca ²⁺ , and Mg ²⁺ and wherein A' is preferably Γ, Br, Cl\ OH ^" , NO3-, HCO3-, HSO4-, and CHsCOO ^" . This allows the anionic portion of the Venturello catalyst to be separated as a salt of M' or in an alternative form when the peroxophosphotungsten moiety cannot retain its structural integrity due to the highly basic conditions. [0052] In Figure la, the general scheme for catalyst retention and recovery described above (Figure 1) is modified to show a specified cation exchange resin followed by a specified anion exchange resin. The cation exchange resin used for catalyst retention is Lewatit® TP 207 and the anion exchange resin used for catalyst retention is Lewatit® S 4528.

[0053] In Figure 2, a general scheme for catalyst retention and recovery using an anion exchange resin followed by a cation exchange resin is shown. In certain embodiments, the cation exchange resin used may comprise a quaternary ammonium group. For catalyst retention, the anion exchange resin shown in step 1 may comprise a compound represented by [N(R')3] ⁺ C1 ^" , wherein R' preferably represents an H group, a CFb group, or a CH2CH3 group. For catalyst retention, the cation exchange resin shown in step 2 may comprise a compound represented by XH, wherein X preferably represents an anion of acetic acid, iminodiacetic acid, sulfonic acid, or phosphonic acid. For catalyst recovery shown in step 3, M'A' is used as an eluent in an aqueous solution, wherein M' preferably represents K ⁺ , Na ⁺ , NH4 ⁺ , Ca ²⁺ , or Mg ²⁺ and A' preferably represents Γ, Br, Cl\ OH ^" , N0 ₃ ^" , HCO3 ^" , HS0 ₄ \ or CH3COO-. For catalyst recovery shown in step 4, HA is used as an acid, wherein HA preferably represents CH3COOH, HCOOH, (COOH) ₂ , H02CCH(OH)CH(OH)C0 ₂ H, HCl, HNO3,

[0054] In Figure 2a, the general scheme for catalyst retention and recovery using an anion exchange resin followed by a cation exchange resin described above (Figure 2) is modified to show a specified anion exchange resin followed by a specified cation exchange resin. The anion exchange resin used for catalyst retention is chemically modified Lewatit® S 4528 and the cation exchange resin used for catalyst retention is Lewatit® TP 207.

[0055] In Figure 3, a scheme for retention and removal of a catalyst using a single bed resin column is shown. An anion exchange resin, for example Lewatit® S 4528, comprising an -NR2 group, is modified with HCl, giving an anion exchange resin with an N ⁺ HR2C1 ^" group. A catalyst or a product mixture comprising a catalyst is diluted with toluene or ethanol and passed through the modified anion exchange resin, retaining the catalyst in the anion exchange resin represented as (-N ⁺ HR2)3M, and giving the phase transfer reagent represented as RC1 in the eluate. An aqueous solution of NH4OH or NH4HCO3 is passed through the anion exchange resin, removing the catalyst represented as (NH^M from the anion exchange resin, now with an (-N ⁺ HR.2)30H group. The phase transfer reagent represented as RC1 in the eluate is passed through a cation exchange resin, such as Dowex™ Mac 3, comprising a -COOH group, retaining the phase transfer reagent in the cation exchange resin. AcOH in toluene or in ethanol is passed through the cation exchange resin, removing the phase transfer reagent represented as AcOR from the cation exchange resin. Generally, MR3 represents a Venturello catalyst, wherein M represents a peroxophosphotungstate anion and R represents a phase transfer cation. For this scheme, the anion exchange resin bed is used first, followed by the cation exchange resin bed. A flow rate of around 4 bed volumes (BV, wet)/hour under gravity is maintained. Double bed resin may be used following the same sequence. Alternatively, this scheme is applicable for removing catalyst under a stirring/shaking condition.

[0056] In Figure 4, results for catalyst separation using cation exchange resin followed by anion exchange resin in a mixed bed system, wherein the cation exchange resin is packed on top of the anion exchange resin within a single column, is shown. Overall, 97.5% retention of total tungsten is achieved.

[0057] In Figure 5, results for catalyst separation using cation exchange resin followed by anion exchange resin in a single bed resin system is shown. Overall, 98.14% retention of total tungsten is achieved.

[0058] In Figure 6, a scheme for removal of Venturello catalyst components using anion exchange resin is shown. In the first step, the anion resin is modified to convert tertiary amine functional groups to corresponding quaternary amine groups. In the second step, a solution comprising the Venturello catalyst is contacted with the modified anion resin and the Venturello catalyst along with phase transfer reagent are extracted in the eluent. In this scheme, M = peroxophosphotungstate anion and RC1 = the phase transfer reagent Aliquat® 336 (chloride form).

[0059] In Figure 7, a scheme for the calculation of catalyst absorbed/retained in resin by a gravimetric method/analaysis is shown. The catalyst absorbed/retained may

(i- ^z )

be calculated using the equation: (x) = x 100% , wherein y represents the total mass removed and z represents the total mass of catalyst. [0060] In Figure 8, a breakthrough curve for the absorption/retention of catalyst using the anion exchange resin Purolite® Al 11 S (-NR2H ⁺ C1 ^" form) is shown. The catalyst absorption (percent, with respect to maximum capacity of resin) is plotted on the y-axis and the mass of the eluate (grams) is plotted on the x-axis. The curve plateaus at around 28.62% catalyst absorption with respect to maximum capacity of the resin.

[0061] In Figure 9, results for recovered tungsten from an anion exchange resin bed (Purolite® Al 11 S, -N ⁺ HR2C1 ^" form) are shown. The total recovery of tungsten was found to be 102%

[0062] In Figure 10, results for recovered phosphorous from an anion exchange resin bed (Purolite® Al 11 S, -N ⁺ HR2C1 ^" form) are shown. The total phosphorous recovery (-224%) was erroneous, and the first point in the graph shows a tungsten/phosphorous (W/P) ratio of 10.5 instead of the theoretically expected 23.4.

[0063] In Figure 11, results for recovery of tungsten and phosphorous from an anion exchange resin column (Purolite® Al 11 S, -N ⁺ HR2C1 ^" form) are shown. The extent of catalyst absorption was 92.8%. The extent of tungsten recovery was 39.7%. The phosphorous recovery results were found to be erroneous

[0064] In Figure 12, results for recovery of tungsten and phosphorous from an anion exchange column (Purolite® Al 11 S, -N ⁺ HR2C1 ^" form) are shown. The extent of catalyst absorption was 78.6%. The extent of tungsten recovery was 38%. The phosphorous recovery results were found to be erroneous.

[0065] In Figure 13, results for the extent of tungsten and phosphorous recovery from an anion exchange resin (Purolite® Al 11 S, -N ⁺ HR2C1 ^" form) under stirring condition are shown. Overall, the lower the charge ratio of catalyst: resin, the higher the extent of catalyst recovery.

[0066] In Figure 14, a scheme for conversion of the -N ⁺ HR2C1 ^" form of an anion exchange resin to the -N ⁺ HR20H ^" form using an NH4OH solution is shown. In step 1, the anion exchange resin Purolite® Al 1 IS is chemically modified using HCl. In step 2, the modified anion exchange resin resultant from step 1 is chemically modified with 2N NH4OH. The -N ⁺ HR20H ^" form of the anion exchange resin is obtained after using a solution of NaCl in step 3. [0067] In Figure 15, a resin bed system optimized to facilitate separation of Venturello catalyst from a scaled-up soybean oil epoxidation reaction is shown. Two columns, each 24 inches long with 1400 mL capacity, were used in a series. The anion exchange column was set up first in the series, followed by the cation exchange column. To deliver a feed solution to the columns, a Master flex L/S pump (peristaltic pump) with detachable pump head was used. To check the pressure within the columns, a pressure gauge was set up before the columns in the series. To regulate the eluate flow rate, valves were placed after each column.

[0068] In Figure 16, an additional view of the resin bed system optimized to facilitate separation of Venturello catalyst from a scaled-up soybean oil epoxidation reaction is shown.

[0069] In Figure 17, additional views of the resin bed system optimized to facilitate separation of Venturello catalyst from a scaled-up soybean oil epoxidation reaction are shown.

[0070] In Figure 18, an additional view of the resin bed system optimized to facilitate separation of Venturello catalyst from a scaled-up soybean oil epoxidation reaction is shown.

[0071] In Figure 19, results of calibration of the flow rate of a Master flex L/S pump are shown.

[0072] In Figure 20, results of calibration of the flow rate of a Master flex L/S pump using two columns are shown.

[0073] The present invention is more particularly illustrated by the following non-limiting examples:

[0074] The term single bed resin as used herein means a column packed with a single resin. In certain preferred embodiments, the single bed resin may comprise only cation exchange resin, or alternatively only anion exchange resin.

[0075] The term double bed resin as used herein means a column packed with two kinds of resin. In certain preferred embodiments, the double bed resin may comprise cation exchange resin on top of anion exchange resin, or alternatively anion exchange resin on top of cation exchange resin.

[0076] The term mixed bed resin as used herein is synonymous with double bed resin. [0077] Table 1 : Resins Used in the Examples

Resin Name Description CAS Company

Number

Lewatit® TP Weakly acidic, macroporous cation 135620- LANXESS AG, 207 exchange resin with chelating 93-8 Cologne,

iminodiacetate groups Germany

Lewatit® S Weakly basic, macroporous anion 113114- LANXESS 4528 exchange resin 05-9

Diaion™ WA Weakly basic anion exchange resin 69011-17- Mitsubishi 30 2 Chemical, Japan

Diaion™ WK Weakly acidic cation exchange 9052-45-3 Mitsubishi 60L resin Chemical, Japan

Dowex™ Macroporous, weak acid cation 9052-45-3 The Dow Mac-3 exchange resin Chemical

Company

Dowex™ 99 Strongly acidic cation exchange The Dow Ca resin Chemical

Company

Diaion™ WK Weakly acidic cation exchange 111559- Mitsubishi 100 resin 80-9 Chemical

Diaion™ WK Weakly acidic cation exchange Mitsubishi 40 resin Chemical

Diaion™ Weakly acidic cation exchange 11159-80- Mitsubishi WTO I S resin 9 Chemical

Lewatit® S Weakly acidic cation exchange 130353- LANXESS 8528 resin 60-5

Dowex™ Strongly acidic cation exchange The Dow 99CA/2010 resin Chemical

Company

Purolite® Macroporous, weak base anion 68441-29- Purolite A111 S exchange resin 2 [0078] Examples 1 -4: Catalyst separation using single bed resin column - cation exchange resin followed by anion exchange resin. A solution of toluene and Venturello catalyst, wherein the Venturello catalyst comprises a tri-negatively charged peroxophosphotunsten moiety and a mono-positively charged methyltrioctylammonium cation as a phase transfer reagent (PTR) was processed in a single bed resin column according to the scheme in Figure 1. The solution of toluene and Venturello catalyst was passed through a column packed with the cation exchange resin Lewatit® TP 207. The average flow rate of the eluate was maintained at around 4 bed volumes (BV, wet)/hour (Fig. 1 , step 1). The eluate was collected and passed through a column packed with the anion exchange resin Lewatit® S 4528 (Fig. 1 , step 2). A solution of 1 normal (N) acetic acid (AcOH) in toluene was passed through the cation exchange resin to recover the phase transfer reagent (PTR) as an acetate salt (Fig. 1 , step 3). A solution of IN ammonium hydroxide (NH4OH) was passed through the anion exchange resin to recover the tungsten and phosphorous (Fig. 1, step 4). This procedure was repeated, using the following combinations of cation exchange resins and anion exchange resins: cation exchange resin (Lewatit® TP 207) followed by anion exchange resin (Diaion™ WA 30); cation exchange resin (Diaion™ WK 60L) followed by anion exchange resin (Lewatit® S 4528); cation exchange resin (Dianion™ WK 60L) followed by anion exchange resin (Dianion™ WA 30). While the quantities of reagents varied based on the amounts of resin used, in one specific example, 0.38 g of Venturello catalyst was dissolved in 100 mL of toluene and passed through a single bed resin column packed with cation exchange resin (Lewatit® TP 207). The eluate was collected and passed through a single bed resin column packed with anion exchange resin (Lewatit® S 4528). A solution of IN acetic acid in toluene (100 mL) was passed through the cation exchange resin to recover the PTR as an acetate salt. A solution of IN

NH4OH/NH4HCO3 (100 mL) was passed through the anion exchange resin to recover the tungsten and phosphorous. Table 4, entry 1 represents tungsten and phosphorous recovery results. Table 6, entry 4 represents catalyst retention in the resin bed results.

[0079] Example 5 : Catalyst separation using shaking/stirring condition - anion exchange resin followed by cation exchange resin. The -NR2 group of the anion exchange resin Lewatit® S 4528 was converted to the -N ⁺ R2HC1 ^" group by shaking the anion exchange resin with a solution of IN HC1 (ethanol/water = 1 : 1, volume/volume) for 24 hours (Fig. 2, step 1). The resin was filtered, washed thoroughly with an ethanol/water 1 : 1, volume/volume solution, and air dried. The objective of this chemical modification of the anion exchange resin was to convert the tertiary amine group to the corresponding quaternary amine group. Alternatively, commercially available anion exchange resins with quaternary amine groups may be used. The quaternary amine group is advantageous compared to primary, secondary, or tertiary amine groups for at least two reasons. First, the anionic portion of the Venturello catalyst containing tungsten and phosphorous is more expensive than the phase transfer reagent Aliquat® 336. Therefore, a method to recover the anionic portion of the Venturello catalyst without a required step to recover the Aliquat® 336 (which does not contribute to epoxidized soybean oil decomposition) is cost-effective and less time consuming compared to a method using primary, secondary, or tertiary amine groups. Second, only one column packed with anion exchange resin is used to recover the anionic portion of the Venturello catalyst when using a quaternary amine, and therefore increased recovery of the anionic portion of the Venturello catalyst is expected compared to use of primary, secondary, or tertiary amine groups, which would require use of two successive columns.

[0080] A product mixture comprising an organic phase and a reaction mixture from an epoxidation of soybean oil was diluted with an ethanol/toluene solution (product mixture: ethanol/toluene = 1 :3, volume/volume) and shaken/stirred in the presence of the anion exchange resin containing the -N ⁺ R.2HC1 ^" group for 18 hours (Fig. 2, step 2). The organic phase comprised soybean oil, epoxidized soybean oil, toluene (about 10%), by-products (diols, about 1-3%), and Venturello catalyst (about 1.9% by weight). The product mixture in all examples was from an epoxidation of soybean oil (ADM feedstock) conducted in a 50 mL autoclave reactor under N2 atmosphere in semi- batch mode using Venturello catalyst comprising a tri-negatively charged

peroxophosphotungsten moiety and a mono-positively charged methytrioctylammonium cation as a phase transfer cation, with 50% aqueous H2O2 added slowly with a syringe pump. In all iterations, the product mixture was obtained from a fresh batch and compositionally the same. The anion exchange resin was removed, washed, and shaken in a solution of IN ammonium hydroxide for 20 hours (Fig. 2, step 3). The eluate from the diluted product mixture shaken/stirred in the presence of the anion exchange resin was collected and stirred with the cation exchange resin Dowex™ Mac-3, H ⁺ form, for 20 hours (Fig. 2, step 4) and then stirred in a solution of IN acetic acid in

ethanol/toluene for 20 hours (Fig. 2, step 5). The cation exchange resin was filtered off and the ethanol/toluene was evaporated to yield the phase transfer reagent (PTR)

Aliquat® 336 (trioctylmethylammonium chloride, CAS 5137-55-3, Sigma- Aldrich) as a viscous liquid (identified by proton (1H) NMR).

[0081] Example 6: Catalyst separation using single bed resin column - anion exchange resin followed by cation exchange resin. The -NR2 group of the anion exchange resin Lewatit® S 4528 was converted to the -N ⁺ R2HC1 ^" group by shaking the anion exchange resin with a solution of IN HC1 (ethanol/water = 1 : 1, volume/volume) for 24 hours (Fig. 3, step 1). The resin was filtered, washed thoroughly with an ethanol/water 1 : 1, volume/volume solution, and air dried. A product mixture comprising an organic phase [soybean oil, epoxidized soybean oil, toluene (about 10%), by- products (diols, about 1-3%), Venturello catalyst [P0 ₄ {WO(0 ₂ )2}4]

[{(C8Hn)3N(CH3)} 3] (about 1.9% by weight)] was diluted with an ethanol/toluene solution (product mixture: ethanol/toluene = 1 :3, volume/volume) and passed through a column packed with the anion exchange resin containing the -N ⁺ R2HC1 ^" group (Fig. 3, step 2). A basic solution (IN ammonium hydroxide or ammonium bicarbonate) was passed through the anion exchange resin to recover the peroxophosphotungstate anion as an ammonium salt (Fig. 3, step 3). The eluate from step 2 was collected and passed through a column packed with the cation exchange resin Dowex™ Mac-3, containing the -COOH group (Fig. 3, step 4). A solution of 1 normal (N) acetic acid (AcOH) in toluene or ethanol was passed through the cation exchange resin to recover the phase transfer reagent (PTR) as an acetate salt (Fig. 3, step 5).

[0082] Example 7: Catalyst retention and recovery. A heptane solution of the Venturello catalyst (1.93 g of Venturello catalyst in 300 mL of heptane) was passed through a resin column at a flow rate of about 35 mL/hour under gravity. Both a mixed bed resin column (double bed resin column) and a single bed anion exchange resin column were tested. For the mixed bed resin system, a cation exchange resin was packed at the top of the column and an anion exchange resin was packed at the bottom of the column. For the single bed resin system, the eluent was passed through the cation exchange resin first, followed by the anion exchange resin. The Venturello catalyst contained 28.42% tungsten and 1.44% phosphorous.

[0083] A solution of p-toluene sulfonic acid in acetonitrile was passed through the resin column but catalyst recovery was negligible (0.54%). A solution of acetic acid (3.1 equivalents with respect to tungsten) was passed through the resin column but did not improve catalyst recovery.

[0084] A solution of ammonium hydroxide and a solution of ammonium bicarbonate were passed through the resin column separately to recover tungsten either as NH4WO4 (water-soluble) or as an ammonium salt of the phosphotungsten moiety. A total of 100 mL of either solution (ammonium hydroxide, ammonium bicarbonate) was passed through the resin column at a flow rate of about 15 mL/hour (about 4 bed volumes (BV)/hour).

[0085] Regarding the solution of ammonium bicarbonate (IN), results for recovery of the catalyst from both the single bed anion exchange resin and the mixed bed resin system were compared. Around 43% of the tungsten and around 94% of the phosphorous was recovered from the mixed bed resin system. However, only around 8.1 % of the tungsten and around 16.7% of the phosphorous was recovered from two successive single bed anion resins.

[0086] Regarding the solution of ammonium hydroxide, results for the recovery of the catalyst from both the single bed anion exchange resin and the mixed bed resin were compared using different concentrations of the solution of ammonium hydroxide. Around 8% of the tungsten and around 4.4% of the phosphorous was recovered from the single bed anion exchange resin when using IN ammonium hydroxide. Around 15.8% of the tungsten and around 21.5% of the phosphorous was recovered from the single bed anion exchange resin when using 8N ammonium hydroxide. When the amount of catalyst was doubled in the heptane solution before passing through the resin column, the recovery of the tungsten and the phosphorous using a solution of ammonium hydroxide (8N) decreased to about 12.1 % of the tungsten and 10% of the phosphorous. When the amount of resin was doubled in the column, the recovery of the tungsten and phosphorous using a solution of IN ammonium hydroxide decreased to about 4.6% of the tungsten and about 2.4% of the phosphorous. [0087] The details of the results obtained from Example 7 are summarized below in Table 2.

[0088] Table 2: Evaluation of Catalyst Recovery (Tungsten and Phosphorous)

Varying Different Parameters

[0089] Example 8: Catalyst recovery. Solutions of ammonium hydroxide and ammonium bicarbonate were used to recover tungsten and phosphorous from columns packed with a mixed bed resin system and a single bed resin. The flow rate was maintained at around 4 bed volumes (BV)/hour for the single bed resin column and around 2 BV/hour for the mixed bed resin system.

[0090] The recovery of catalyst (tungsten and phosphorous) was compared varying the amounts of resins in the mixed bed resin column. The total amount of resin was tested using 4 grams (g), 20g, and 80g, with the amount of cation exchange resin and anion exchange resin present in equal amounts. The amount of the tungsten recovered increased from about 41% when 4g of resin was used, to about 61% when 20g of resin was used, to about 100% when 80g of resin was used. The Venturello catalyst contained 28.42% tungsten and 1.44% phosphorous.

[0091] The details of the results obtained from this portion of Example 8 are summarized below in Table 3.

[0092] Table 3: Evaluation of Catalyst Recovery (Tungsten and Phosphorous)

Using Mixed Bed Resin with Varied Amounts of Resin

[0093] The recovery of catalyst (tungsten and phosphorous) was compared for different anion exchange resins. To improve recovery of the tungsten and phosphorous, the resin in the column was removed after passing the eluent through the resin, stirred with the eluent for about 4 to about 5 hours, filtered and washed with deionized water. This additional stirring with the eluent yielded an additional recovery of about 9% to about 25% of the tungsten and an additional recovery of about 12% to about 28% of the phosphorous. The numbers reported in parentheses in Table 3 below represent the total amount of the element collected after removing the resin from the column after elution and stirring the resin with 50 mL of a IN solution of ammonium bicarbonate for 40 hours followed by washing with deionized water.

[0094] The details of the results obtained from this portion of Example 8 are summarized below in Table 4.

[0095] Table 4: Evaluation of Catalyst Recovery (Tungsten and Phosphorous) Using Mixed Bed Resin with Different Types of Anion Resin

[0096] The eluent was passed through the mixed bed twice to improve recovery of the tungsten and the phosphorous. The resin beds were activated with a solution of acetic acid in toluene. The Venturello catalyst contained 28.42% tungsten and 1.44% phosphorous. The details of the results obtained from this portion of Example 8 are summarized below in Table 5.

[0097] Table 5: Evaluation of Catalyst Recovery (Tungsten and Phosphorous) Passing the Eluent through the Mixed Bed Resin System Twice

[0098] The retention of catalyst (tungsten and phosphorous) in the resin bed was compared. For the mixed bed resin column, the cation resin Lewatit® TP 207 yielded better retention than the cation resin Dianion™ WK 60L. Passing the solution containing the catalyst through the same column multiple times increased the catalyst retention. The solvent used was toluene. The Venturello catalyst contained 28.42% tungsten and 1.44% phosphorous. To retain the catalyst, 100 mL of the catalyst solution in toluene was passed through the resin.

[0099] The details of the results obtained from this portion of Example 8 are summarized below in Table 6 and Table 7. [00100] Table 6: Evaluation of Catalyst Retention (Tungsten and Phosphorous)

Varying Combinations of Resin in Single Bed Resin and Mixed Bed Resin Systems

[00101] Table 7: Evaluation of Catalyst Retention (Tungsten and Phosphorous) Reusing the Mixed Bed Resin System

Entry Column Amount Amount Type of Eluent to Recovered Recovered

Type (m of resin of resin remove tungsten phosphorus

= mixed used catalyst 'W and (weight (weight bed (g) 'P' prior to %) %) resin) reuse

1 m 20 1.93 Lewatit® - 99.9 99.9

1 ^st TP 207

use (cation),

Lewatit®

S 4528

(anion)

2 m 20 1.93 Lewatit® IN NH4OH 98.5 98.3

ReTP 207 (100 mL)

use (cation),

Lewatit®

S 4528

(anion)

(anion)

[00102] Example 9: Catalyst and Phase Transfer Reagent (PTR) Recovery. Solutions containing catalyst and phase transfer reagent were shaken with different resins and eluents for several hours. Anion exchange resin was used first, followed by cation exchange resin. The -NR2 group of the anion exchange resin Lewatit® S 4528 was converted to the -NR2HCI group by shaking the resin with a 50mL solution of IN HC1 (methanol/ water = 1 : 1) for 24 hours. The modified anion exchange resin was filtered off and washed with a 1 : 1 methanol/water solution. A solution of 40 mL of acetonitrile and lg Venturello catalyst was shaken with the modified anion exchange resin for 25 hours. The modified anion exchange resin was filtered off and washed with a 1 : 1 methanol/water solution. The PTR was Aliquat® 336. A total of lg of the Venturello catalyst was used in each entry, along with 8g of Lewatit® S 4528. The Venturello catalyst contained 28.42% tungsten and 1.44% phosphorous.

[00103] The amount of phase transfer reagent (PTR) was greater than 100%. The details of the results obtained from this portion of Example 9 are summarized below in Table 8.

[00104] Table 8: Evaluation of Phase Transfer Reagent (PTR) Recovery Using Anion Exchange Resin

[00105] Resin was shaken in various eluents with 80mL of a methanol/water 1 : 1 (volume/volume) solution for 40 hours to recover the catalyst (tungsten and phosphorous). The eluents tested were IN ammonium bicarbonate, IN sodium hydroxide, and IN ammonium hydroxide. The three eluents had similar efficiency for recovering the tungsten. A total of lg of catalyst was used in each entry, along with 8g of Lewatit® S 4528. The catalyst was loaded with 28.42% tungsten and 1.44% phosphorous. The retention of the catalyst in the resin bed was considered to be 100%.

[00106] The details of the results obtained from this portion of Example 9 are summarized below in Table 9.

[00107] Table 9: Evaluation of Neat Catalyst (Tungsten and Phosphorous)

Recovery Using Anion Exchange Resin and Different Eluents

[00108] Example 10: Catalyst Recovery Using Solutions Containing Catalyst and a Product Mixture. An isolated product mixture containing epoxidized soybean oil (ESO) and catalyst was shaken with the anion exchange resin Lewatit® S 4528 and various eluents for 22 hours to determine amounts of tungsten and phosphorous recovered. A total of lOg of epoxidized soybean oil containing 0.27g of catalyst was used in each entry, along with 8g of Lewatit® S 4528. The catalyst was loaded with 28.42% tungsten and 1.44% phosphorous. The retention of the catalyst in the resin bed was considered to be 100%.

[00109] The details of the results obtained from Example 10 are summarized below in Table 10.

[00110] Table 10: Evaluation of Catalyst (Tungsten and Phosphorous)

Recovery Using a Product Mixture, Anion Exchange Resin, and Different Eluents

[00111] Example 1 1 : Phase Transfer Reagent (PTR) Recovery. Various microporous weak acid cation exchange resins were screened to compare recovery of the phase transfer reagent (PTR) Aliquat® 336 under shaking conditions. Improved retention and recovery of the PTR was observed using a macroporous weak acid resin (Dowex™ Mac-3) and a microporous strong acid resin (Dowex™ 99 CA, H ⁺ form). These resins which provided improved retention and recovery of the PTR were also tested by packing in columns.

[00112] The details of the results obtained from Example 11 are summarized below in Table 11 and Table 12. For Table 10, 6g of dry cation exchange resin was shaken with solutions of PTR in a mechanical shaker for 24 hours. To retain the PTR, 0.4g of Aliquat® (the PTR) in 20 mL of toluene and 30 mL of acetonitrile was used (entries 1 -4). To recover the PTR, 20 mL of 2N acetic acid in toluene was used. For Tables 11 and 12, for the shaking conditions, 6g of dry cation exchange resin was shaken with solutions of PTR in a mechanical shaker for 24 hours. To retain the PTR, 0.4g of Aliquat® (PTR) in 20 mL of toluene was used. To recover the PTR, 20 mL of 2N acetic acid in toluene was used. For Table 12, for the resin column, 6g of dry cation exchange resin was packed in the column. A solution containing 0.4g of Aliquat® (PTR) in 50 mL of toluene was eluted at a flow rate of 10 mL/hour to retain the PTR in the resin bed. A 50 mL solution of 2N acetic acid was eluted at a flow rate of 10 mL/hour to recover the PTR.

[00113] Table 1 1 : Evaluation of Phase Transfer Reagent (PTR) Retention and Recovery Using Different Cation Exchange Resins under Shaking Conditions

Entry Cation exchange resin Retained PTR (wt. Recovered PTR (%), w.r.t

%) retained PTR

1 Diaion™ WK100 (NH ₄ ⁺ 8.0

form)

2 Diaion™ WK100 (Na ⁺ 11.7

form)

3 Lewatit® TP 207 (NH ₄ ⁺ 11.1

form)

4 Lewatit® TP 207 (Na ⁺ 11.9

form)

5 Dianion™ WK40 29.9 25.1

a orm

[00114] Table 12: Evaluation of Phase Transfer Reagent (PTR) Retention and Recovery Using Different Cation Exchange Resins under Shaking Conditions and Resin Column

[00115] Example 12: Phase Transfer Reagent (PTR) Recovery Using Solutions Containing PTR. Catalyst, and a Product Mixture. A product mixture containing epoxidized soybean oil (ESO), catalyst, and PTR was passed through a resin bed to retain the catalyst, followed by passing an ethanol/ammonium hydroxide solution through the resin bed to recover the catalyst. Using Lewatit® S 4528 as the anion exchange resin and Dowex™ Mac 3, H+ form, as the cation exchange resin, 71.3% of the PTR (Aliquat® 336) was recovered from the single bed resin column. Using the same reagents but substituting shaking conditions instead of using the single bed resin columns, 73.3% of the PTR (Aliquat® 336) was recovered. For the single bed resin column entry, 12.3g of epoxidized soybean oil (ESO, ER-74) in 50 mL of ethanol was passed through 5g of Lewatit® S 4528, HCl salt form, at a flow rate of 15 mL/minute, and the eluate was then passed through 5g of Dowex™ Mac-3, H+ form, and then eluted with a solution of IN acetic acid in 50 mL of ethanol at a flow rate of about 15 mL/minute. Calculations were done considering 100% retention of the catalyst in the resin bed. [00116] The details of the results obtained from Example 12 are summarized below in Table 13.

[00117] Table 13: Evaluation of Phase Transfer Reagent (PTR) Recovery Using a Product Mixture, Shaking Conditions, and a Single Bed Resin Column

[00118] Example 13: Catalyst separation using cation exchange resin followed by anion exchange resin. Both single bed resin and mixed bed resin systems were evaluated. For the single bed resin system, the cation exchange resin was used first, followed by the anion exchange resin. For the mixed bed resin system, the cation exchange resin was packed at the top of the column and the anion exchange resin was packed in the lower part of the column. For all of the columns, 5g of epoxidized soybean oil product mixture was loaded onto lOg of resin (cation resin = Lewatit® TP 207; anion resin = Lewatit® S 4528), with a wet bed volume of about 17 mL. For the single bed resin system, the flow rate was about 45 mL/hour (about 2.5 bed volumes (BV)/hour). For the mixed bed resin system, the flow rate was about 40 mL/hour. Toluene was used as the eluent (about 100 to about 130 mL).

[00119] The results of Example 13 are shown in Fig. 4 and Fig. 5. For the mixed bed resin system, retention of total tungsten was 97.5%. For the single bed resin system, retention of total tungsten was 98.14%.

[00120] Examples 14-18: Absorption capacity of anion exchange resin under stirring condition. In order to determine the optimal charge ratio of catalyst to resin (catalystresin) to get maximum retention of the Venturello catalyst in the resin, five experiments were done using a fixed amount of dry resin (3.5 g) and different equivalents of charge of catalyst (catalystresin ranging from 1 : 1 to 1 : 12). A

macroporous, weak base anion exchange resin (Purolite® Al 11 S) was used for all experiments. The resin was modified to convert the tertiary amine functional groups to the corresponding quaternary amine group. (Figure 6). The resin was stirred with a 40g Venturello catalyst solution in toluene at a mild agitation speed (300 rpm) for 12 hours. After 12 hours of stirring, 4g of solution was removed and dried. Gravimetric analysis was performed to obtain qualitative results. (Figure 7). The total amount of catalyst retained was calculated from the isolated mass of catalyst and/or catalyst components. Results, summarized below in Table 14, showed that the extent of Venturello catalyst absorbed varied dramatically with charge ratio.

[00121] Table 14: Qualitative Estimation of Catalyst Absorbed/Retained Using Purolite® Al 11 S Anion Exchange Resin by Gravimetric Method, Stirring Condition

[00122] Example 19: Breakthrough point of anion exchange resin used for Venturello catalyst absorption. The breakthrough point corresponds to how many times a resin column may be used to absorb catalyst before saturation. In order to determine the breakthrough point, a single bed resin column was filled with 3.5 g of dry Purolite® Al l I S anion exchange resin. Twelve catalyst solutions were prepared using 0.5 g of Venturello catalyst and 19.5 g of toluene (20g catalyst solution, total) used as eluent. The resin column contained a charge equivalent to 6 g of Venturello catalyst

(catalys resin charge ratio = 1 : 12), with the wet volume of resin totaling about 7 mL. Each catalyst solution was passed through the resin column with an average eluate flow rate of 15 mL/hour (2 BV/hour) under gravity. The content of catalyst or catalyst components was analyzed from each set of eluate (20 g) by gravimetric method to estimate the extent of catalyst absorption in each run. It was expected that at some point, there would be no catalyst absorption, which would correspond to the breakthrough point. The results are summarized below in Table 15. After the 6 ^th run onward, catalyst absorption dropped significantly. After the 11 ^th run onward, no Venturello catalyst absorption was noted. Therefore, the resin bed became saturated after the 11 ^th run and the maximum catalyst absorption at the breakthrough point was 28.62% with respect to the catalyst absorption measured by gravimetric method. These results are summarized in Table 16 below and Figure 8.

[00123] Table 15: Qualitative Estimation of Catalyst Absorbed/Retained Using Purolite® Al 11 S (-NR2 HCI form) Anion Exchange Resin by Gravimetric Method, Resin Column

[00124] Table 16: Qualitative Estimation of Catalyst Absorbed/Retained Using Purolite® Al 11 S Anion Exchange Resin by Gravimetric Method, Resin Column

Run 1 2 3 4 5 6 7 8 9 10 11 12

ID E63- E63- E63- E63- E63- E63- E63- E63- E63- E63- E63- E63- 1 2 3 4 5 6 7 8 9 10 11 12

Isolated 0.27 0.35 0.40 0.42 0.43 0.41 0.47 0.48 0.48 0.47 0.496 0.503 mass (g)

Catalyst 99.4 64.8 43.2 34.6 30.2 38.9 13.0 4.3 4.3 13.0 1.7 0 absorbed

in each

run (%) Total 8.23 13.63 16.96 19.84 22.36 25.6 26.68 27.04 27.4 28.48 28.62 28.62 catalyst

(%)

absorbed

w.r.t.

charge

[00125] Examples 20-24: Optimization of the catalvs resin charge ratio for effective absorption of Venturello catalyst. In order to determine the optimal charge ratio of catalyst: resin to obtain maximum retention of Venturello catalyst, as well as to determine the maximum absorption capacity of resin for a particular catalyst: resin charge ratio, five different experiments were conducted using a fixed amount of dry resin (3.5 g) and different equivalents of charge ratio of catalyst: resin (from 1 :3 to 1 : 12). A 20g solution of Venturello catalyst in toluene was prepared using different amounts of catalyst for the different catalys resin charge ratios. Each catalyst solution was passed through a resin column at an average eluate flow rate of 15 mL/hour (2 BV/hour) under gravity. The content of catalyst or catalyst components was analyzed for each set of eluate (20 g) by gravimetric method to estimate the extent of catalyst absorption. Using the gravimetric method, total catalyst absorbed/retained was calculated from the isolated mass of catalyst and/or catalyst components. The results are summarized below in Table 17. The higher the charge ratio of catalyst: resin, the greater the extent of catalyst absorption/retention. Based on these results, a catalyst: resin charge ratio of about 1 : 12 may be considered as an optimum ratio for effective absorption of Venturello catalyst for a one-time run.

[00126] Table 17: Optimization of Catalyst: Resin Charge Ratio for Effective Absorption/Retention of Catalyst, One Time Run, Gravimetric Method

Entry ID Catalyst Charge of Isolated Catalyst used (g) catalyst: resin mass (g) absorbed (%)

1 E68- 0.5 1 : 12 0.272 98.5

1

E63- 0.5 1 : 12 0.270 99.4

1

2 E64- 1.1 1 :6 0.63 92.3

1 3 E64- 1.65 1 :4 1.10 72.0 2

4 E64- 2.2 1 :3 1.56 62.8

3

[00127] Examples 25-27: Optimization of eluate flow rate for effective absorption of Venturello catalyst. The flow rate of eluate plays an important role in effective absorption/retention of catalyst. Generally, the greater the flow rate, the lower the absorption capacity. From literature, a flow rate of about 2-3 BV (wet/swelled)/hour was determined to be desirable. To validate this for the present invention, three separate catalyst separation experiments were conducted using 0.5 g Venturello catalyst in 34.5 g toluene and a catalyst: resin ratio of 1 : 12. The catalyst solutions were passed through the resin bed with different flow rates. The applied eluate flow rates were about 14 mL/min (2 BV/hour), about 21 mL/min (3 BV/hour), and 28 mL/min (4 BV/hour), respectively. Based on gravimetric estimation, the lowest flow rate (2 BV/hour) gave the maximum absorption of Venturello catalyst (98.5%) whereas the highest flow rate (4 BV/hour) gave the lowest absorption (78.6%) and the intermediate flow rate (3 BV/hour) gave mediocre absorption (92.8%). The results are summarized below in Figure 18.

Therefore, an eluate flow rate of about 2-3 BV/hour may be considered an optimum flow rate.

[00128] Table 18: Optimization of Eluate Flow Rate for Effective Absorption of

Catalyst, Anion Exchange Resin Bed [Purolite® Al 1 IS, -N ⁺ HR2C1 ^" form], Gravimetric Method

Entry ID Flow rate Catalyst Charge of Isolated Catalyst of eluate used (g) catalyst: mass (g) absorbed resin (%)

1 E68- ~14 mL/h 0.5 1 : 12 0.272 98.5

1 (2 BV/h)

2 E68- -21 mL/h 0.5 1 : 12 0.285 92.8

2 (3 BV/h)

3 E68- -28 mL/h 0.5 1 : 12 1.318 78.6

3 (4 BV/h) [00129] Examples 28-31 : Combined effect of catalyst concentration and eluate flow rate on catalyst absorption capacity using anion exchange resin. The combined effect of catalyst concentration and eluate flow rate was studied in four experiments. In the first two experiments (Table 19, entries 1 and 2), catalyst concentration was varied (2.5% by weight and 1.4% by weight), while the eluate flow rate was maintained at 2 BV (swelled)/hour. In the last experiments (Table 19, entries 3 and 4), the eluate flow rate was varied (2 BV/hour and 3 BV/hour), while the catalyst concentration was maintained at 1.4% by weight. From these results, it was determined that almost similar absorption of catalyst could be achieved (-99%) by varying catalyst concentration from 2.5% by weight to 1.4% by weight, while the extent of catalyst absorption decreased slightly (from 98.5% to 92.8%) when the eluate flow rate was increased from 2 BV/hour to 3 BV/hour. If the catalyst concentration is decreased further (to 1% by weight) while the eluate flow rate is increased to 4 BV/hour, the extent of catalyst absorption decreases further (from 92.8% to 78.6%). Under current scaled-up epoxidation reaction conditions, the organic layer contains about 2.6% by weight catalyst. Therefore, the reaction mixture containing catalyst may be diluted 2-3 times for effective absorption of catalyst.

[00130] Table 19: Combined Effect of Catalyst Concentration and Eluate Flow Rate on Catalyst Absorption/Retention Capacity of Anion Exchange Resin, Gravimetric Method

Entry ID Flow Mass of Catalyst Isolated Catalyst rate of solution concentration mass (g) absorbed eluate (g) (wt. %) (%)

1 E63- -14 20 2.5 0.270 99.4 1 mL/h

(2 BV/h)

2 E68- -14 35 1.4 0.272 98.5 1 mL/h

(2 BV/h)

3 E68- -21 35 1.4 0.285 92.8 2 mL/h

(3 BV/h)

4 E69- -21 50 1.0 1.318 78.6 2 mL/h

(3 BV/h) [00131] Example 32: Determination of Extent of Catalyst Recovery from Saturated Resin Column. In this experiment, catalyst was removed from a saturated resin column used for the breakthrough experiments. The extent of catalyst recovered was calculated with respect to the absorbed/retained catalyst measured by the gravimetric method. Samples of eluate were collected, warmed to expel NH3 from residual NH4OH, and analyzed by ICP. After eluting 280 g of IN aqueous NH4OH solution, the total recovery of tungsten was found to be 102%. (Figure 9).

[00132] Conditions for obtaining the saturated resin column were: 3.5 g dry Purolite® Al l I S anion exchange resin; 0.5 g Venturello catalyst dissolved in 19.5 g toluene (20 g total solution) as eluent; 0.5 g of catalyst in each run (total of 12 runs); catalyst: resin charge ratio = 1 : 12; wet volume of resin = approximately 7 mL; and eluate flow rate = approximately 15 mL/minute (2 BV/hour). The total absorption was 28.62% (at breakthrough point) with respect to the total charge of resin, and 26.95% with respect to the total mass of catalyst run (6 g).

[00133] Conditions for catalyst removal/recovery from the saturated column were: IN NH4OH as eluent; flow rate = 3 BV/hour (wet resin); 280 g eluent total;

sodium tungstate dehydrate Na2W04-2H20) as ICP standard.

[00134] The first 35 g of eluate contained the maximum extent (-98%) of tungsten, meaning that only a small amount of eluent is required for maximum recovery of catalyst when using a resin column saturated with catalyst. The phosphorous recovery results showed a trend similar to that of tungsten, but the total recovery (-224%) was erroneous, and the first point in Figure 10 shows a tungsten/phosphorous (W/P) ratio of 10.5 instead of the theoretically expected 23.4. (Figure 10). Therefore, it would be preferred to dilute the eluate (especially the first amount of eluate obtained) before using the eluate for ICP analysis.

[00135] Examples 33 and 34: Determination of Extent of Catalyst Recovery from an Unsaturated Resin Column. Two similar experiments were conducted in which 0.5 g of Venturello catalyst was used for a catalyst absorption study (catalyst: resin charge ratio = 1 : 12). The extent of catalyst absorption was 92.8% (Figure 1 1) and 78.6% (Figure 12). The extent of tungsten recovery was 39.7% (Figure 11) and 38% (Figure 12). The phosphorous recovery results were found to be erroneous. [00136] The conditions for the experiment represented by Figure 11 (Example

33) were: IN NH4OH as eluent; flow rate = 3 BV/hour (wet resin) (21 mL/hour); 57 mL eluent total; sodium tungstate dehydrate (Na2W04-2H20) as ICP standard. Total absorption of catalyst was 92.8% with respect to the total catalyst used (0.5 g Venturello catalyst, absorbed/retained using 35 g toluene at a flow rate of 3 BV/hour).

[00137] The conditions for the experiment represented by Figure 12 (Example

34) were: IN NH4OH as eluent; flow rate = 3 BV/hour (wet resin) (21 mL/hour); 57 mL eluent total; sodium tungstate dehydrate (Na2W04-2H20) as ICP standard. Total absorption of catalyst was 78.6% with respect to the total catalyst used (0.5 g Venturello catalyst, absorbed/retained using 35 g toluene at a flow rate of 4 BV/hour).

[00138] Examples 35-39: Determination of Extent of Catalyst Recovery from Resin under Stirring Condition. Five different experiments were conducted using a fixed amount of dry resin (3.5 g, Purolite® Al 11 S anion exchange resin, -N ⁺ HR2C1 ^" form) and different amounts of catalyst (catalyst: resin ratio varying from 1 : 12 to 1 : 1). The resin was stirred in NH4OH (40 g, aqueous solution) for 12 hours at an agitation speed of 300 rpm. The loading of tungsten and phosphorous in the catalyst is 32.6% and 1.374%, respectively. For each experiment, a slight white precipitate was noticed at the bottom of the samples, and therefore erroneous results may have been obtained as the concentration may have been outside the ICP range. The extent of recovered tungsten is summarized in Table 20 and Figure 13. The tungsten recovery of 67% for catalys resin ratio of 1 : 1 seemed to be erroneous as the concentration of this sample was out of the ICP range. Overall, the lower the charge ratio of catalyst: resin, the higher the extent of catalyst recovery.

[00139] Table 20: Recovery of Tungsten from Anion Exchange Resin under Stirring Condition.

Entr ID Catalys Catalyst:Resi Catalyst Tungsten Tungsten Tungsten y t Used n Charge Absorbe Absorbe Recovered Recovered

(g) Ratio d (%) d (mg) (mg) (%)

1 E62 0.5 1 :12 95.0 154.8 67.4 43.5

-1

2 E62 1 1 :6 92.8 302.5 141.8 46.9

-2 3 E62 2 1 :3 54.0 352.1 253.2 71.9

-3

4 E62 4 1 :1.5 51.3 668.9 529.4 79.1

-4

5 E62 6 1 :1 41.4 809.8 543.4 (may 67.1 (may be

-5 be erroneous erroneous because because concentratio concentratio n out of ICP n out of ICP range) range)

[00140] Example 40: Effect of Chemical Modification of Purolite® Al 11 S Anion Exchange Resin. The use of anion exchange resin with -N ⁺ HR20H ^" form instead of -N ⁺ HR2C1 ^" form can possibly reduce the acid content in epoxidized soybean oil. Therefore, in this experiment, the N ⁺ HR20H ^" form was attempted to be prepared from the N ⁺ HR2C1 ^" form using a 2N aqueous NH4OH solution. Use of a relatively strong base, such as NH4OH, creates a high chance of formation of the -NR2 form instead of the -N ⁺ HR ₂ OH- form. (Figure 14). In this experiment, 30 g of Purolite® Al 11 S (- N ⁺ HR2C1 ^" form) was stirred in 50 mL of a 2N aqueous NH4OH solution for 3 hours. The resin was washed with deionized water several times, followed by a solution of acetone and water, and finally with acetone. After filtration, the resin was dried. Approximately 23 g of resin was collected. Reduction of weight of the resin indirectly proved formation of the -NR2 form, the -N ⁺ HR20H ^" form, or both. The resin (5 g) was stirred with deionized water (50 mL) for 30 minutes. After filtration, the filtrate was preserved in order to check for hydroxyl content. Addition of phenolphthalein solution did not show any pink coloration, indicating that the filtrate did not contain any OH ^" groups that might have come from residual NH4OH. The washed resin was stirred in 50 mL of an aqueous NaCl (4 g) solution for 6 hours. The resin was filtered off and the filtrate was collected to check for hydroxyl content. The filtrate was titrated with a 0.68 N aqueous HC1 solution (titrated by a standard NaOH solution). The measured hydroxyl content was 41 milli-equivalents per kilogram (at least 20 times less than commercially available resin). Therefore, this process was not effective for converting anion exchange resin with the -N ⁺ HR ₂ C1- form to the -N ⁺ HR ₂ OH- form. [00141] Example 41 : Large scale separation of Venturello catalyst. In order to separate the Venturello catalyst form the scaled-up (518 g soybean oil) epoxidation reaction, an adequate resin bed system was optimized. Two columns 24 inches long with 1400 mL capacity were used. The anion exchange column was set up first, followed by the cation exchange column. To deliver the feed, a Master flex L/S pump (peristaltic pump) with detachable pump head was used. To check the pressure within the columns, a pressure gauge was set up before the columns. To regulate the eluate flow rate, valves were placed after each column. Figures 15-18 show the set up. The flow rate was calibrated using water as the solvent. (Figures 19-20).

[00142] After the scaled-up epoxidation of soybean oil, about 500 g of toluene and 500 g of double-distilled water were added to the reaction mixture and stirred for 1 minute to quench the reaction and facilitate the separation of the aqueous and organic phases. After 30 minutes, the organic phase was separated from the aqueous phase and treated with anhydrous Na2S04 (50 g) to remove the residual H2O/H2O2 from the organic phase. The organic phase was further diluted with toluene (from about 350 g to about 500 g) to reduce the viscosity.

[00143] The resins used were Purolite Al 1 Is (-NR2FLCI ^" form) [anion exchange resin] and Dowex Mac 3 (H ⁺ form) [cation exchange resin]. The anion exchange resin column was placed first, followed by the cation exchange resin column. Each resin column was filled with 480 g of dry resin in an adequate amount of toluene. The wet volume of the anion and cation exchange resin columns were 960 mL and 580 mL, respectively. The anion exchange resin column was prepared using 500 mL of toluene, whereas the cation exchange resin column was prepared using 400 mL of toluene and 700 mL of acetone. These resin columns were used to separate Venturello catalyst (7.6 g in each batch) from the scaled-up (518 g soybean oil) epoxidation reaction mixtures of three different batches. The diluted reaction mixture (1620 g) of the first batch (ER-102) was run through the columns one time at a flow rate of 20 mL/min (about 1.2-2 BV/hour). The diluted reaction mixtures (1420 g) of the second (ER-103) and third (ER-104) batches were run through the resin columns two times each with set flow rates of 20 mL/min (1 ^st run) and 30 mL/min (2 ^nd run). In order to collect residual product from inside the resin column, 400 mL of toluene was passed through each column at the end of the catalyst separation for each batch. The total charges of resin (480 g) in each column compared to the catalyst present (about 7.6 g) in the reaction mixture of each batch was about 70: 1. The results are summarized in Table 21 and Table 22 below.

[00144] Table 21: Details about Purification of Epoxidized Soybean Oil (ESO) using Ion Exchange Resin Column - Resin Column Set-up

[00145] Table 22: Details about Purification of Epoxidized Soybean Oil (ESO) Ion Exchange Resin Column - Product Purification

Product Amount Toluene Anhydrous Set flow rate Toluene (to be used (to Na2S04 used (mL/minute) used at purified) dilute) (to remove end of run residual water)

ER-102 Approximately 1620 g 50 g 20 400 mL

580 g

ER-103 Approximately 1420 g 50 g 1 ^st run: 20 400 mL

580 g 2 ^nd run: 30

ER-104 Approximately 1420 g 50 g 1 ^st run: 20 400 mL

580 g 2 ^nd run: 30 [00146] In order to remove toluene from the epoxdized soybean oil product, the diluted product was roto-evaporated two times for about 40-45 minutes at 70°C and about 40 minutes at 80°C, respectively. After the first time, the color of the product remained similar to the color of the diluted product. However, after the second time, the color of the product noticeably changed.

[00147] The present invention has been described with reference to certain examples. However, it will be recognized by those of ordinary skill in the art that various substitutions, modifications, or combinations of any of the examples may be made without departing from the spirit and scope of the invention. Thus, the invention is not limited by the description of the examples, but rather by the appended claims as originally filed.